Re: entailment-from-inconsistent-graph: new test case request

>Summary: I doubt you can make clearer tests of this issue than the one peter
>proposes - in particular Jan's proposed test is incorrect.
>
>
>while sympathetic to Jan's desire for meangingful test cases, I note that
>peter's test (particualrly the choice of conclusion) depends on a somewhat
>non-obvious technicality.
>
>false does-not-simply-entail G
>
>for any non-empty G, because G uses at least one URI which is not in all
>interpretations of false.

No, it does. A entails B means (forall interpretations I (if I 
satisfies A then I satisfies B))

so if A is false then this is always trivially true; so false entails 
anything, no matter what vocabulary it uses.

I removed the vocabulary restriction on entailment at Herman's 
insistence: he was right.

Pat

>Thus the test case needs to show an inconsistency, followed by garbage using
>the same vocab as the premise (or  the central RDF vocabulary). (And we have
>rdf-entailment), ditto rdfs entailment).
>
>Peter's choice of garbage out
>
>rdf:type rdf:type rdf:type
>
>seems to be about as best as you can do to have obvious garabge given the
>constraint
>
>Jeremy
>
>
>


-- 
---------------------------------------------------------------------
IHMC	(850)434 8903 or (650)494 3973   home
40 South Alcaniz St.	(850)202 4416   office
Pensacola			(850)202 4440   fax
FL 32501			(850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes

Received on Friday, 7 November 2003 10:11:01 UTC